Back Up Next

Begin planisphere module  - February 2018

This month Buchanan begins the second to the last mechanical complication on the machine, planisphere. The last will be the orrery.



The first two photos are the initial axis design for the nesting of the star plate and sun hand indicator. The third is an initial gear layout. Notice how Buchanan uses paper cutouts for the gear arrangements. The fourth photo shows the initial gearing for the sun hand precession. In this example the hand remains stationary when the plastic disk below (not easily seen in this photo) is rotated. This was made to demonstrate what later will have a set of gears that will allow the hand to have an annual precession across the sky. In other words, the star plate rotates once in a sidereal day with the sun rotating once in a solar day. A sidereal day is about  3 minutes, 55.9 seconds shorter than a solar day thus after one solar day the sun hand will be just under four minutes ahead of the star plate. So if on January 1st both the sun hand and a point marked on the star plate are aligned at twelve o’clock, the sun hand will advance just under four minutes each day and after one year the sun hand will return to its twelve o’clock position aligned with the mark on the star plate. This is the representation of the sun’s precession across the sky as it relates to its position against the stars. See video below.



The first photo shows the gear set needed to drive the planisphere from the movement. The drive is a sidereal input for the star plate. The second photo is a print out from the on line compound gear calculator, a real labor saving device from the sources we used initially in 2006 for gear calculations.

Buchanan writes: I have tried a second sidereal gear train to the first. The first had gears that were too small to cover the distance between demo output and the planisphere so I worked out a larger set but these would not fit into the available space so I am now looking for an intermediate size gear train. I have found a workable gear train. There are a few clock parts in the way, here or there. 1 second in 336,000 years should be acceptable. I have a 12 gear train that is 1 second in 24 million years. I would have liked a more arty set but the combinations do not oblige. Tomorrow I can fill in all the final details.

I had a three day run on the timing machine and there are good temperature cycles each day but no sign of any temperature error at all!

This last sentence is good news for the accuracy of the clock; the Elinvar springs are doing their job.


Buchanan writes: In the attached photos, the pens are pointing to the edge of the bezel next to the winding squares. The first photo with the small gears is the sidereal train and the other, with the big gear is the time train. This will be in front of the sidereal train. So you will see from the front of the clock 7 gears and 8 will be hidden. The two faint circles at top left and top right are the 24 hour dial and the thermometer.

Buchanan writes: I have the design complete and everything is fitting on well (after a little juggling). I have had to add one more large wheel in to push a gear train backward so as to be able to fit in the support frame.  It will give us a little visual complexity. I will ink In the drawing now and then I can start cutting metal tomorrow.

Buchanan writes: I have all the gear sizes worked out so am now preparing blanks, I hope to start cutting gears on Monday, I may possibly first start the plinth and support frame for two reasons, firstly I have lost one of my gear cutters and am waiting for another from England. And secondly I need to stress relieve the brass sheet for the gears before I start.  

This is the final drawing. Buchanan is now ready to cut metal.


Buchanan writes: I have all my blanks roughed out and stress relieved.  I hope to have them all machined to size by tonight. Some are old gears that I am repurposing; I have cut doubles sometimes but never seem to use them. I am also working on the big blocks for the base, Brass on the left and plastic on the right in photo.


Buchanan writes:  I have the disc for mounting the enamel on roughed out, and, a method thought out as how to mount the enamel onto the brass dial plate. I also have the two main bearing sleeves complete with one gear attached.  Next is the collet and centre shaft for the dial plate and then the main frame for the planisphere. Then the process of mounting the drive trains and floating them on as much air as possible.



Buchanan writes: I have the main frame with the centre bearing fitted. The clearance’s for the Fasoldt’s have caused a little trouble and some juggling  but all is working out well now. The next requirement is the attachment to the main frames. I will have to mount the base and mainframe properly before I go any further because everything else depends on the fixed position of the centre shaft. The Fasoldts Buchanan refers to are the detent whips mounted to the strike train fan assemblies.


Here we see Buchanan has been able to move the planisphere disk forward enough to clear the Fasoldt strike train detent whips, see video below. But another problem has surfaced. The main barrel winding arbors are too close to the planisphere disk. They just barely clear the disk and any misalignment of the user inserting the winding key for the quarter strike train will cause damage to the edge of the planisphere bezel. The diameter cannot be reduced since the enamel star disk has already been made. The winding arbor on the other side for the celestial train presents the same problem. It looks like these will need to be extended forward so the key will fit before it reaches the vertical plane where the planisphere dial bezel is located.




Buchanan writes: I have the pillar holes drilled in the main frame.  I had to make a special jig and a custom clamp.   It feels like sacrilege to drill into a running clock with a 3/8 inch drill! The two stainless steel pillars are almost complete. I will have a lever each side of the dial, but inside, to lock this frame in place. The pedestal and bezel will mount on this main frame and come away as a complete unit.  

This is truly an example of “measure twice cut once”. Here Buchanan secures the planisphere support frame to the clock base rail. If a mistake is made here, the main clock frame is ruined! Fortunately everything went to plan. It’s interesting how sometimes a critical fabrication step devolves into the use of a simple home drill.


Buchanan writes: I have the spiral bayonet fittings almost complete. The spiral on each pillar is opposite to the other. So by lifting the clamp levers on each side the planisphere is released and by pushing both clamp levers down it clamps both pillars. I must make recesses in the frames for the clamp levers and the pillar screws and the clamps will be finished.  Then I can start on the gear train.



Deryck writes: The clamps are almost complete. The machining was a little tricky as there is a raised ridge around the pillar hole that runs in the groove in the clamp arm so as to retain it in the recess to make installing easy.


These two photos show the planisphere module mounting bracket secured to the clock base frame from the outside and inside.



The first photo shows the male and female components of the levers that allow the user to quickly mount or dismount the planisphere module. Next a close up of one of the levers.


These two photos show the pair of levers that secure the planisphere to the clock frame. The photos show the unlocked and then locked positions.


This wheel is the drive for the planisphere disk. It is one of the largest yet finest wheels in the machine with 516 teeth at just under 5” (12 cm) in diameter. Look at the delicacy of this wheel; it could easily be crushed in one hand.

The wheel mounted to the planisphere frame. It is one of the largest and most delicate examples of Buchanan's wheel cutting artistry. It's unfortunate that this beautiful piece of wheel work art will be largely hidden forever after the planisphere enamel dial and bezel is installed, so look at it now! 



These photos show the sun hand drive. One can see some conflict where the center of the drive wheel is butting up against the rear plate rim of the thermometer module in the last photo.


The rear plate of the thermometer module shown in the first photo is further skeletonized in the second. This was needed for both esthetic as well as the need for clearance for the sun hand drive.




These photos show the build out of the thermometer unit and the sun hand drive. The thermometer unit was never a stand-alone module and was always mounted to a frame that held an intermediate wheel connecting the demonstration crank to the planisphere through the sun hand drive as well as the tellurian.

Back Up Next